For the cropped image, Why do we Merge AB into M and CD into N? Also, where do those numbers in the calculation(cropped image) come from?

Specifically, the (1-5), (1-6), & (5-6)? I need to understand it.

ADM3308-Winter 2020: Business Data Mining Press esc to exit full screen Q6) Consider the following data set Data = { , } (a) Cluster the data using agglomerative clustering technique with single linkage. Show the similarity (distance) matrix at each step. Use Manhattan distance function. Let A = , and D = Similarity Matrix X-X+ y - y1 Calculations Note: Distance from x to y = Distance from y tox AB=11-21 + 1-2-|-1| + |-1 - 2 BC=12-51 + 2-21 = 3 + 10 = 3 C --D-15-61 +12-10-|-1| + |-1 - 2 D 5 5 B 0 2 5 5 B 2. 0 3 5 0 2 2 0 D Smallest Distance = 2 Merge (AB) into M. Note: {CD) can be merged into another cluster (N) at the same time since. Calculations C-M = MIN((1-5 + 11-21.12-5 + 12-20) = MIN (5,3)= 3 D- + M = MIN((1-6] + |1-11, 12-6 + 12-1D) - MIN (5,5) = 4 C --D-15-6 + 12-11-1-1+11 - 2 M 0 M D C 3 0 2 D 5 2. 0 5 University of Ottawa Telfer School of Management Page 1 of 3 Merge {A B) into M.* Note: {CD} can be merged into another cluster (N) at the same time since. Calculations C++ M = MIN((1-51 + |1-2, 12-51 + 12-2D) = MIN (5, 3) = 3 D M = MIN((1-6] + |1-11, 12-61 + 12-1)) = MIN (5,5) = 4 C--D=15-6] + |2-1 = |-1| + |11 = 2 M D M 0 3 5 3 0 2 D 5 2 0 ADM3308-Winter 2020: Business Data Mining Press esc to exit full screen Q6) Consider the following data set Data = { , } (a) Cluster the data using agglomerative clustering technique with single linkage. Show the similarity (distance) matrix at each step. Use Manhattan distance function. Let A = , and D = Similarity Matrix X-X+ y - y1 Calculations Note: Distance from x to y = Distance from y tox AB=11-21 + 1-2-|-1| + |-1 - 2 BC=12-51 + 2-21 = 3 + 10 = 3 C --D-15-61 +12-10-|-1| + |-1 - 2 D 5 5 B 0 2 5 5 B 2. 0 3 5 0 2 2 0 D Smallest Distance = 2 Merge (AB) into M. Note: {CD) can be merged into another cluster (N) at the same time since. Calculations C-M = MIN((1-5 + 11-21.12-5 + 12-20) = MIN (5,3)= 3 D- + M = MIN((1-6] + |1-11, 12-6 + 12-1D) - MIN (5,5) = 4 C --D-15-6 + 12-11-1-1+11 - 2 M 0 M D C 3 0 2 D 5 2. 0 5 University of Ottawa Telfer School of Management Page 1 of 3 Merge {A B) into M.* Note: {CD} can be merged into another cluster (N) at the same time since. Calculations C++ M = MIN((1-51 + |1-2, 12-51 + 12-2D) = MIN (5, 3) = 3 D M = MIN((1-6] + |1-11, 12-61 + 12-1)) = MIN (5,5) = 4 C--D=15-6] + |2-1 = |-1| + |11 = 2 M D M 0 3 5 3 0 2 D 5 2 0